This type of information sharing method between neighboring nodes is used in order to increase follow-up capabilities with respect to a topology change and an information change in an upper layer while reducing the amount of messages for information sharing in the upper layer by sharing information with neighboring nodes in the upper layer, triggered by the detection event of changes in the neighboring state between nodes and in topology in a lower layer of a network.
For example, a case will be described in which a routing protocol is used to exchange topology information in the lower layer, and at the same time, information such as node information and positional information is exchanged in the upper layer.
When messages are exchanged independently in each layer, a message exchange is generally performed in the lower layer at short update intervals, therefore, follow-up with respect to changes in the neighboring state and topology is fast. In contrast to this, in the upper layer, since a message exchange is performed at relatively long update intervals, follow-up with respect to changes in the neighboring state and topology is relatively slow.
Therefore, instead of performing message exchanges independently in each later, a message exchange for information sharing in the upper layer is performed, triggered by the detection of a topology change in the lower layer, allowing follow-up with respect to the topology change in the upper layer to be accelerated.
For example, Patent Document 1 describes a network management system in which a line in an upper layer affected by failure and performance degradation is controlled, triggered by the detection of failure or performance degradation of a line in a lower layer. Specifically, Patent Document 1 describes such examples as performing control to bypass a corresponding link in the upper layer.
In addition, Patent Document 2 describes a network node device detecting a change in topology map information, which contains information on the topology in a layer 2, and updating a neighbor registration table in a layer 3 to regenerate a routing table in the layer 3, triggered by the detection of the update of the neighbor registration table.
Patent Document 1: Japanese Patent Laid-Open No. 2002-354038
Patent Document 2: Japanese Patent Laid-Open No. 2006-157716
The information sharing method between neighboring nodes described in the related art has the following problems.
A first problem is that, following a topology change in the lower layer, even information that is unnecessary to update in the upper layer is exchanged, leading to an increase in the amount of message exchanges in the upper layer.
Every time a topology change is detected in the lower layer, a message exchange for information sharing is performed in the upper layer, therefore, for example, even if a sufficient number/ratio of neighboring nodes having information on the node in question are present because a message for information sharing has been transmitted until the last time, a message exchange for information sharing is performed each time, leading to heavy loads.
A second problem is that the load of a message exchange for information sharing in the upper layer becomes considerably high if the topology is changed frequently in the lower layer.
Every time a topology change is detected in the lower layer, a message exchange for information sharing is performed in the upper layer, therefore, if the topology is changed frequently, even for the information for which frequent update is not necessary in the upper layer because the update frequency is not high, a message exchange for information sharing is performed each time, leading to heavy loads.
It is an exemplary object of the present invention to provide an information sharing method between neighboring nodes, a node and an information sharing program capable of reducing a load caused by a message for information sharing in an upper layer.